The circadian clock is an internal timing mechanism that exists to link a number of biological processes with the environmental light-dark cycle. Because these clocks are inherently related to night and day, photic stimulation is a logical (and existent) input into vertebrate clocks.

Clock inputs such as the light-dark cycle establish the phase and period of the pacemaker (Moore 1997). Eliminating an input via lesion, ablation, or occluding causes a free-running period in the corresponding clock. For example, blinded rats are unable to reset their corticosterone levels via the night-dark cycle (Wilson et al. 1976), possibly yielding loss of food anticipatory activity (Diaz-Munoz et al. 2000). Studies involving clock inputs have generally focused upon photic entrainment pathways, wherein photic information travels via the retinohypothalamic tract (RHT) to the suprachiasmatic nucleus (SCN) (Sutin 1993; Bae 1998; Suri 1998). Sectioning the RHT in hamsters causes the SCN to display a free-running period, demonstrating that the clock runs independently of its photic input. This would suggest the necessity of photic cues in the entrainment of the SCN (Jonson et al. 1988). Nonetheless, numerous studies have demonstrated that clocks of hamsters and mice can synchronize (entrain) to such inputs as a running wheel (Kas et al. 2001), treadmill (Mistlberger 1991), arousal by a saline or triazolam injection (Van Reeth et al. 1989), changes in gravity (Murakami 2000), feeding activity (Stokkan 2001), and numerous other influences other than a light-dark schedule. In 1984, Rosenwasser et al. demonstrated that rhythmic oscillations in food anticipatory activity continued in the absence of feeding. Furthermore, there appeared to exist varied coupling activity between the two clocks (feeding and photic entrained) that depended upon the feeding intervals (Rosenwasser 1984). Later studies on rats (White and Timberlake 1994) supported Rosenwasser’s (1984) conclusions by showing that rats that normally anticipated feedings with increased locomotor activity lost such anticipation in the dark after being fed during the inactive day or given free access to food at all times.

Restricted feeding entrains components of the digestive system

Food anticipatory activities in rats include a rise in core temperature, elevated serum corticosterone, and an increase in duodenal disaccharidases (Stephan 2002). As these behaviors generally involve digestive organs, experiments have monitored feeding entrainment in the pancreas (Muller et al. 1985), intestines (Comperatore et al. 1987), and the liver (Hara et al. 2001). Inversed phases in Dbp mRNA (see Albrecht 2002 for a complete description) expression for mice fed during the day versus night also support the conclusion that restricted feeding can entrain expression in peripheral tissue (Damiola et al. 2000).

Of all the peripheral organs, the liver was the quickest to entrain (Damiola et al. 2000). The results of the experiment imply that the liver may serve a central function in the food-entrained oscillator, thus necessitating future studies involving the food-entrained properties of the liver and its relationship to the photic entrained SCN.

Entrainment of the liver is independent of the SCN

While White and Timberlake (1994) demonstrated coupling between the food-entrained and photic-entrained oscillators, the specific nature of that coupling remains unknown. Understanding how feeding entrainment in the liver relates to photic entrainment in the SCN is essential in understanding how these components are able to effectively couple to one another. Conflicting data from White and Timberlake (1994) seem to indicate the relationship between photic food entrainment is more complex than the current model wherein the SCN serves as a master pacemaker.

Stokkan et al. (2001) demonstrated an increase in pre-meal locomotor activity associated with feeding entrainment of the liver and SCN (Figure 1), a result consistent with studies indicating that restricted feeding does not entrain multi-neuronal activity in the SCN (Inouye 1982). While liver oscillators may couple to the SCN via rhythmic feeding, results (Stokkan et al. 2001) suggest that such feeding entrains the liver independently of the SCN.

Figure 1 . Graph A depicts bioluminescence in a randomly fed rat. Graph B depicts a rat on a four-hour restricted feeding regimen over seven days. Both rats were exposed to identical 12/12 LD cycles. Arrows indicate times livers were explanted. Tissue phase is reflected by the peak of the first subjective day. Comparison of A and B indicates significant phase shift in one day and increase in bioluminescence before feeding (Stokkan et al. 2001). The SCN remained phase-locked to the light cycle while the liver shifted its rhythm significantly (in accordance with the rapid entrainment found by Damiola et al. 2000). Graphs from Stokkan et al. (2001).

In 2001, Hara et al. demonstrated that the SCN, under free feeding conditions, is necessary for circadian rhythms of liver mPer1 and mPer2 expression, but that restricted feeding-induced oscillation of mPer1 and mPer2 in the liver occurs independently of the SCN. These results support the cooperative nature of the SCN and liver oscillators.

Perone et al. (2003) demonstrated that Clk/Clk mutant mice displayed food anticipatory activity in both LD (light-dark) and DD (dark-dark) schedules. While mutants were arrhythmic in DD for SCN-related locomotor activity rhythms, they displayed increased FAA in DD as compared to LD. Consequently, the temporal cue provided by the food signal either exhibited a reduced effect, or light masked the FEO (Perone et al. 2003). While both oscillators work together in some manner, each can operate independently of the other (Perone et al. 2003).

Stress-induced responses to restricted feeding may affect the FEO

Results obtained by Stokkan et al. (2001) suggest that corticosterone mediated stress does not affect phase rhythmicity in the liver. Balsalobre et al. (2000) used an in vivo assay to shift Dbp rhythms using dexamethasone. Accordingly, in vivo assays with the glucocorticoids might demonstrate a corticosterone stress-related response. Further experimentation will be required to determine whether the phase shifts are a result of a property of the food itself or simply the stress associated with food restriction.

How the FEO operates remain a mystery

The specific molecular mechanisms of the FEO remain unknown. If neural or hormonal entraining signals from the SCN to peripheral oscillators exist, they may not have been measurable/visible to Stokkan et al. (2001) due to food entrainment of the FEO in the liver (and other organs) and the photic input to the SCN entraining feeding behavior. Perone et al. (2003), furthermore, imply that restricted feeding entrains peripheral oscillators without affecting the SCN.

While the input pathway for the FEO remains unknown, several feasible possibilities exist. For example, daytime feeding in mice is known to elicit a depression in body temperature during the night (Damiola et al. 2000), demonstrating the ability of clocks to entrain to temperature cycles. Such alterations in temperature may act as the entraining mechanism in the FEO. If holding internal body temperature constant does not impact the feeding-entrained rhythms, temperature fluctuations are not likely to be required inputs.

Pinpointing the location of the FEO

While the liver has been implicated as the principal FEO (Stokkan et al. 2001), future experiments involving the ablation of other potential candidates will be required for confirmation. Davidson et al. (2001) used bulbectomized and sham-operated rats to rule out the olfactory bulb as a factor involved in the food response.

The PBN, a region in the brain responsible for integrating information from visceral and gustatory afferents, may play a role in initiating FAA. Davidson et al. (2000) found that the PBN is involved as a relay center or possible output in the food-entrainment pathway. However, the determined the area is not likely the FEO.

Neural centers responsible for food-related responses, such as the fornix located beneath the corpus callosum, should be studied in the future as a potential site for the FEO. Rats with fornix transections exhibit increased frequencies of eating as well as other differential eating behaviors (Osbourne 1986). Determining whether or not feeding-entrainment is possible with a fornix transection will aid in determining whether the area is necessary for food-related responses. Experiments demonstrating the ability of the fornix to entrain to restricted feeding in the presence of SCN ablation might also suggest sufficiency.

Stokkan et al. (2001) implicated the liver as the possible FEO. While the SCN controls rhythms in gene expression in the liver, this control seems to shut down during restricted feeding schedules. The mechanism of this uncoupling of the SCN from the liver remains unknown. Data collected in restricted feeding assays seem to imply that food plays a role in decoupling the SCN from the liver. However, it is unknown if food quality (taste, smell, chemical composition, texture, nutrients, etc.) or other effects caused by food ingestion (stress) yields such coupling.

Although Dbp and Rev-erbα mRNA appear to be molecules controlled by the oscillator in the liver (Balsabore et al. 2000), it is unknown if rats in which the Dbp and Rev-erbα genes have been knocked out still entrain to restricted feeding cycles. If such knockouts display free running periods in FAA, these genes may be important inputs. If the knockouts are arrhythmic, the genes are likely involved in the oscillator or output pathway.

Due to its key role in the metabolism of food products, the liver will likely serve as the subject of future experiments involving food-entrained oscillators in vertebrates. An experiment monitoring the effects of food-entrainment on liver DBP cycling in SCN-lesioned animals is necessary to confirm the truly independent (if so) nature of DBP as a component of this peripheral oscillator. If SCN lesioning results in a loss of rhythmicity of such cycling, then DBP cannot be considered entirely independent from the SCN. On the other hand, if such rhythms persist in the absence of a functioning SCN, DBP may in fact play a key role in maintaining an independent rhythm based on the availability of food products.

Addressing these questions will not only help pinpoint the location of the food entrained oscillator but also aid in a more thorough understanding of the molecular mechanisms behind the entrainment and how they relate to the light entrained oscillator.